Method, device, medium and system for structured information generation for orbital images
By performing coarse-to-fine segmentation and feature fusion on the original orbital image, high-precision and fast structured information of the orbital image is generated, which solves the problems of subjective interference in manual interpretation and time-consuming and labor-intensive segmentation models, and improves the accuracy and efficiency of structured information.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING TONGREN HOSPITAL AFFILIATED TO CAPITAL MEDICAL UNIV
- Filing Date
- 2026-04-14
- Publication Date
- 2026-07-17
AI Technical Summary
In existing technologies, the generation of structured information by manually interpreting multi-sequence orbital images is easily affected by subjective factors, resulting in inaccurate information and omission of key features. Furthermore, high-precision segmentation models are time-consuming while fast segmentation models lack sufficient accuracy, making it difficult to balance segmentation accuracy and speed.
The first segmented image is obtained by coarsely segmenting the original orbital image, and a spatial attention weight map is generated. This map is then used to perform fine segmentation on the original orbital image. The features of the original orbital image, the labeled image, and the segmented image are combined to generate structured information.
It achieves high-precision and fast orbital image segmentation, ensuring the accuracy of structured information and the efficiency of generation, and resolving the contradiction between segmentation accuracy and speed.
Smart Images
Figure CN122417320A_ABST